[whatwg] Doctype-triggered behavior and the <label> element

[Ian Hickson]

Oops, missed this e-mail.

On Sat, 21 Aug 2004, Matthew Raymond wrote:
> Ian Hickson wrote:
> > On Sat, 31 Jul 2004, Matthew Thomas wrote:
> > > On 31 Jul, 2004, at 7:00 AM, Matthew Raymond wrote:
> > > >
> > > > I recommend that, for HTML documents with the [WF2] doctype, there
> > > > should always be a default option selected, regardless if that default
> > > > is specified.
> > > 
> > > I like that idea.
> > 
> > I strongly disagree with adding more DOCTYPE-triggered behaviour changes.
> > Quirks mode is enough trouble as it is. We don't need yet-more code paths.
> > It makes testing hard. :-)
> 
> How, then, do you justify changing the behavior of <label> with regards 
> to passing events to the associated control? Clearly, with this change 
> in behavior, browsers rendering documents that use <label> must 
> determine the type of the document in order to be both HTML 4.01 and WF2 
> compliant.

It is my interpretation of the two specs that they are not mutually 
contradictory.

However:

> Are you suggesting that user agents that are compliant with 
> WF2 should not be required to be compliant with HTML 4.01?

Yes. WF2 changes several things, WA1 a few more. For example, the default 
"media" for <link> and <style> elements in HTML4 is "screen", and in WA1 
it is "all".


> For that matter, aren't we deliberately forcing UA vendors to choose 
> between doctype-triggered behavior and HTML 4.01 nonconformance when the 
> doctype identifies the document as specifically being HTML 4.01?

Not for the <label> case, but for some other cases, yes. However, in 
almost all of these cases (for example the "media" one above), the vendors 
have _already_ decided to forego HTML4 compliance, without WHATWG even 
existing.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'